INDEX

Explanations

what

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Datuak

-1.00

ValueStyle

-0.88

enumi

-0.84

 الحره

-0.82

 Exactly

-0.81

 بيها

-0.79

DoubleQuotes

-0.78

httphttps

-0.77

setVerticalGroup

-0.74

 Need

-0.71

POSITIVE LOGITS

 conditional

0.35

aile

0.34

ahora

0.33

Regs

0.33

=−

0.32

cierge

0.32

 suppressor

0.32

 deletes

0.31

loroethene

0.31

 Pologne

0.31

Activations Density 0.006%